• swarm intelligence
• swarm
• stigmergy
• flocking
• UAV

The usage of UAVs (Unmanned Aerial Vehicles) has been increasing in the last few years. One of the most important domain applications is the area surveying, i.e. to extract important information from such area. A single UAV equipped with sensors on board is capable of exploring only a limited area because of limited lifetime battery. The usage of multiple UAVs may solve the problem but introduces some coordination issues.
The aim of this Thesis is to develop a simulator of swarm UAVs coordinated via stigmergy and flocking emergent mechanisms and to adapt the swarm in different scenarios, as illegal dumps, forest fire, urban mine and so on. The stigmergy is a bio-inspired mechanism used by ants and bees and it is used as indirect and distributed communication way among drones. The flocking is a behaviour exhibited by birds flying in groups and it is used as a work-group paradigm in order to locate potential targets and to increase the sensing reliability. These two mechanisms can be adapted depending on the scenario, finding the optimal combination of related parameters via an evolutionary algorithm. The simulator of swarm UAVs has been evaluated on some real-world and synthetic scenarios in order to assess the effectiveness of the approach.